Sentry vs GlitchTip
GlitchTip is the closest thing to a free Sentry: open source, compatible with the Sentry client SDKs, and light enough to self-host. Here's exactly when it wins and what you give up.
The short version
If your reason for leaving Sentry is cost or wanting your data in-house — and you don't want to rewrite instrumentation — GlitchTip is the natural move. You keep your existing Sentry SDKs and point their DSN at GlitchTip instead of sentry.io. If your reason is that you want more (deep tracing, session replay, an AI debugger), GlitchTip is the wrong direction; stay on Sentry.
How they compare
| Sentry | GlitchTip | |
|---|---|---|
| License | Proprietary (+ OSS self-host) | Open source (MIT) |
| SDK | Sentry SDKs | Sentry SDKs (compatible) |
| Free tier | 5k errors/mo | 1k events/mo hosted · unlimited self-host |
| Self-host | Yes — 40+ containers | Yes — a handful of containers |
| Paid from | ~$26/mo (Team, annual) | $5/mo hosted · server cost self-host |
| Tracing / spans | Deep | Basic performance |
| Session replay | Yes | No |
| AI debugging (Seer) | Yes ($40/contributor) | No |
| Uptime / cron checks | Yes (metered) | Yes |
When GlitchTip wins
- You want to escape Sentry's bill without touching application code — same SDKs, new DSN.
- You need data on your own infrastructure but can't justify running self-hosted Sentry's 40+ containers.
- You mostly need error tracking, not a full observability suite.
When to stay on Sentry
- You rely on deep tracing, session replay, or Seer AI — GlitchTip doesn't match these.
- You want a fully managed vendor and your volume fits inside Sentry's quotas.
- Your team would rather not own any self-hosting, even a light one, and 1k hosted events/mo isn't enough.
Migration effort
Low. Because GlitchTip accepts Sentry SDK events, migrating is usually a config change: stand up GlitchTip (hosted or self-hosted), create a project, and swap the DSN. Historical events don't transfer — run both in parallel briefly, then cut over. Whichever you land on, wire it into your pipeline so each deploy tags a release automatically (for example via Buddy's Sentry action, which also targets Sentry-compatible endpoints).